Molecular profiling of colorectal tumors stratified by the histological tumor-stroma ratio

Background: The tumor microenvironment or tumor stroma is a dominant determinant of cancer cell behaviour and disease progression. The presence of high tumor stroma content is likely associated with cancer cells acquiring pro-metastatic capacities. The tumor-stroma ratio (TSR) is a robust prognostic tool that is scored on haematoxylin and eosin (H&E) paraffin sections at the invasive margin of the tumor. The importance of the tumor stroma was also emphasized in the colorectal cancer consensus molecular subtypes (CMS). The CMS classification described four CRC subtypes, of which the poor-prognosis CMS4 subtype is characterized by high stromal infiltration and mesenchymal gene expression. However, the biological mechanism of the tumor stroma is not well understood. We therefore aim to investigate the overall transcriptomic profiles and activated pathways of tumors classified by the TSR method using gene expression data. Methods: Seventy-one patients with stage I-III colorectal cancer were included in the study with available gene expression data and H&E sections. Firstly, we scored the TSR on H&E sections and performed survival analysis. Secondly, we quantified the amount of stromal cells present in the tumor bulk of the gene expression analysis.Thirdly, we investigated the biological pathways differently activated between the two groups using online currated gene sets of the MSigDB. Finally, we compared the association between the TSR classification and the CMS classification assessed based on gene expression data. Results: The TSR was an excellent prognostic marker in a multivariable analysis [OS p = 0,0001, HR = 4,59 (1,96 – 10,75)]. Tumors stratified as stroma-high based on histological TSR had significantly more stromal cells compared to stroma-low tumors (p = 2,58*10^-5). Pathways related to epithelial-mesenchymal-transition, angiogenesis, extracellular matrix and integrin were significantly different in the two tumor-stroma groups. We found that the TSR and the CMS classification were associated (χ2 test = 7,71; p = 0,005). Conclusions: At the moment, we are validating our findings in a second cohort and further investigating genes present the relevant pathways. These results will be presented on the day of the symposium. Legal entity responsible for the study: Wilma Mesker, assistant professor LUMC Funding: None Disclosure: All authors have declared no conflicts of interest.